Artificial Intelligence and Industrial Carbon Emissions in Spain: A Fixed-Effects and Mediation Analysis of Structural Transformation
AUTHORS
Lucía Martínez-Vega,Department of Energy Engineering, Universidad de Sevilla, Seville, Spain
Javier Torres-Sánchez,School of Industrial and Telecommunications Engineering, Universidad de Cantabria, Santander, Spain
Elena Ruiz-Navarro,Department of Computer Science and Artificial Intelligence, Universidad de Granada, Granada, Spain
ABSTRACT
The decarbonization of industrial systems is a central engineering challenge for achieving climate-neutrality targets across Europe, particularly in Spain, where energy-intensive industries continue to contribute substantially to national carbon emissions. In parallel, the rapid advancement of Artificial Intelligence (AI) and Industry 4.0 technologies presents new opportunities to optimize industrial processes and improve energy efficiency. However, empirical evidence on the extent and mechanisms through which AI influences industrial carbon emissions in the Spanish context remains limited. This study investigates the impact of AI on industrial carbon emissions in Spain by employing a panel-data analysis framework that integrates fixed- and mediated-effects models. Drawing on regional-level data, AI development is proxied using patent-based indicators, while industrial carbon emissions are estimated using energy consumption and standardized emission coefficients. The analysis examines both the direct effects of AI adoption on emissions and the indirect effects mediated through industrial structural transformation, specifically the advancement and rationalization of industrial structure. The results indicate that AI adoption significantly reduces industrial carbon emissions, demonstrating robustness across multiple model specifications. Furthermore, mediation analysis reveals that AI contributes to emission reductions by promoting the transition to advanced industrial structures and improving industry-wide resource allocation efficiency. The indirect effects account for a substantial share of the total impact, underscoring the importance of structural transformation pathways for achieving sustainable industrial outcomes. These findings provide important engineering and policy implications. From a systems engineering perspective, AI-enabled optimization and intelligent process control can serve as critical levers for reducing industrial carbon intensity. From a policy standpoint, fostering AI innovation alongside targeted industrial restructuring strategies can accelerate Spain's transition toward a low-carbon economy. Overall, this study contributes to the growing literature on digitalization and sustainability by offering a context-specific, empirically grounded analysis of the AI–carbon emissions nexus in Spain.
KEYWORDS
Artificial intelligence, Industrial carbon emissions, Industrial structure transformation, Low-carbon manufacturing, Spain
REFERENCES
[1] European Commission, "A Clean Planet for All: A European Strategic Long-Term Vision for a Prosperous, Modern, Competitive and Climate Neutral Economy," Brussels (2019)
[2] D. García-Gusano and D. Iribarren, “Prospective energy security scenarios in Spain: The future role of renewable power generation technologies and climate change implications,” Renewable Energy, vol. 126, pp. 202–209, (2018). DOI:10.1016/j.renene.2018.03.044(CrossRef)(Google Scholar)
[3] S. H. Ivanov and C. Webster, “Adoption of robots, artificial intelligence and service automation by travel, tourism and hospitality companies – A cost-benefit analysis,” Presented at the International Scientific Conference “Contemporary Tourism – Traditions and Innovations,” Sofia University, (2017). Available: https://ssrn.com/abstract=3007577
[4] T. Zheng, M. Ardolino, A. Bacchetti, and M. Perona, “The applications of Industry 4.0 technologies in manufacturing context: A systematic literature review,” International Journal of Production Research, vol. 59, no. 6, pp. 1922–1954, (2021). DOI:10.1080/00207543.2020.1824085(CrossRef)(Google Scholar)
[5] Spanish Government, “Spain Digital Agenda 2025,” Madrid, (2020)
[6] H. Borgi, N. Alessa, F. Hamza, and K. Albitar, “The impact of green innovation and renewable energy on CO2 emissions in G7 nations,” Heliyon, vol. 10, no. 10, Art. no. e31142, (2024). DOI:10.1016/j.heliyon.2024.e31142(CrossRef)(Google Scholar)
[7] Y. Lu, “Industry 4.0: A survey on technologies, applications and open research issues,” Journal of Industrial Information Integration, vol. 6, pp. 1–10, (2017). DOI:10.1016/j.jii.2017.04.005(CrossRef)(Google Scholar)
[8] P. X. Hoa, V. N. Xuan, and N. T. P. Thu, “Factors affecting carbon dioxide emissions for sustainable development goals – New insights into six Asian developed countries,” Heliyon, vol. 10, no. 21, Art. no. e39943, (2024). DOI:10.1016/j.heliyon.2024.e39943(CrossRef)(Google Scholar)
[9] M. Dong, G. Wang, and X. Han, “Impacts of artificial intelligence on carbon emissions in China: In terms of artificial intelligence type and regional differences,” Sustainable Cities and Society, vol. 113, Art. no. 105682, (2024). DOI:10.1016/j.scs.2024.105682(CrossRef)(Google Scholar)
[10] T. K. Lepasepp and W. Hurst, “A systematic literature review of Industry 4.0 technologies within medical device manufacturing,” Future Internet, vol. 13, no. 10, Art. no. 264, (2021). DOI:10.3390/fi13100264(CrossRef)(Google Scholar)
[11] S. Liang, J. Yang, and T. Ding, "Performance evaluation of AI-driven low carbon manufacturing industry in China: An interactive network DEA approach," Computers & Industrial Engineering, vol. 170, Art. no. 108248, (2022). DOI:10.1016/j.cie.2022.108248(CrossRef)(Google Scholar)
[12] A. Parteka and A. Kordalska, “Artificial intelligence and productivity: Global evidence from AI patent and bibliometric data,” Technovation, vol. 125, Art. no. 102764, (2023). DOI:10.1016/j.technovation.2023.102764(CrossRef)(Google Scholar)
[13] IPCC, “2006 Guidelines for National Greenhouse Gas Inventories,” Intergovernmental Panel on Climate Change, (2006)
[14] Z. Wen and B. Ye, “Analyses of mediating effects: The development of methods and models,” Advances in Psychological Science, vol. 22, no. 5, pp. 731–745, (2014)